Knowledge and attitudes of registered nurses towards artificial intelligence at a national hospital in Antigua

Artificial intelligence (AI) is transforming healthcare delivery. Limited evidence exists regarding nurses’ knowledge and attitudes towards AI in the English-speaking Caribbean. We assessed the knowledge, attitudes and perceived competency requirements in AI among registered nurses at a national hospital in Antigua. All 179 registered nurses at the hospital were invited to complete a survey collecting data on demographic characteristics, knowledge of AI, sources of AI-related information and perceptions regarding the AI-related competencies that nurses should possess. Five-point Likert-type questions assessed attitudes toward AI use. Cronbach’s alpha measured internal consistency and ordinal logistic regression tested the significance of independent associations with selected attitudes. Differences in perceived AI competency needs between students, practising and nursing faculty were examined using Cochran’s Q and McNemar tests. Eighty-six nurses (response rate 48.1%, mean age 37.6 years) participated. Only 34.9% indicated awareness of AI’s application in nursing practice. Television/radio (81.4%), social media (80.2%), educational institutions (27.9%) and the workplace (16.3%) were the sources of information on speech to text recognition applications. Most agreed/strongly agreed that AI would revolutionise nursing (52.9%); that automated visit documentation (67%), intervention recommendations (56.0%) problem list generation (55.3%), care intervention identification (50.6%), social determinants of health assessment (50.5%), care coordination recommendations (50.6%) were useful applications; and that nursing education should include information on AI applications (61.2%). Most respondents disagreed/strongly disagreed that AI would soon replace nurses (64.3%) and were neutral about AI making nursing more exciting (63.5%). Cronbach’s alpha for the18 items on attitude towards AI was 0.891. Nurses with diploma/associate degrees were less likely than those with bachelor’s/master’s degree to believe that automated identification of care interventions would be a useful (OR = 0.27, 95% CI 0.11, 0.64). More respondents expected practising and nursing faculty to be competent in technical domains of AI compared to student nurses (p < 0.016). Most respondents were unaware of AI use in nursing practice but felt it would revolutionize nursing and supported its integration into nursing education. Few had workforce replacement concerns. Targeted AI education is needed. Not applicable.